Search the news
Found 18 articles
tagged #llm — page 1 of 2
TypeSafe AI and OpenRouter Launch Jev Smart Router for LLM Efficiency
TypeSafe AI has officially announced a strategic integration with OpenRouter, introducing the Jev Smart Router to optimize Large Language Model (LLM) calls. By implementing the typesafe/jev-router architecture, the plat…
Alibaba Unveils Qwen3.8-Omni-Flash: Cutting Video Costs by 89%
Alibaba Cloud’s AI division, Tongyi Qianwen, has officially launched Qwen3.8-Omni-Flash, a pioneering multimodal model designed with a primary focus on agent-driven capabilities. This release marks a significant milesto…
Perplexity AI Introduces Four-Tier Effort Settings for Task Optimization
Perplexity has announced a significant update to its Computer product, introducing a granular "effort" adjustment system designed to optimize computational resources. Users can now navigate four distinct levels of reaso…
Xiaomi Open-Sources CocktailASR-1: AI Model Solves Multi-Speaker Audio Issues
The technology giant Xiaomi has officially announced the release and open-sourcing of its industrial-grade target speaker Automatic Speech Recognition (ASR) model, Xiaomi-CocktailASR-1. Designed to address the long-stan…
DeepSeek Harness v0.1.5 Launches with V4.1 Flash Model Integration
The developers behind the DeepSeek ecosystem have announced the release of DeepSeek Harness v0.1.5, a significant update designed to enhance the synergy between the interface and the DeepSeek V4.1 Flash model. This late…
New Research Identifies Critical Thresholds for Stablecoin De-pegging
The Renmin University of China Institute of Financial Technology has published a comprehensive study in the Journal of International Money and Finance regarding the systemic vulnerabilities of stablecoins. By utilizing …
Abliteration.ai Debuts Offensive Cybersecurity Model Based on GLM-5.3
The startup Abliteration.ai has officially announced the launch of its latest large language model, abliterated-model-large-v2. Designed specifically for offensive cybersecurity applications, this new iteration leverage…
Silicon Data Secures $30.5M to Fuel AI Computing Benchmark Economy
Silicon Data, an independent benchmark and verification platform for the AI computing economy, has successfully closed a initial $30.5 million Series A funding round. The investment round was led by the Valor Atreides A…
Meta Open-Sources Muse Glimmer: A 30B Agent Model Optimized for Consumer GPUs
Meta has officially announced the release of Muse Glimmer, a 30-billion parameter autonomous agent model, under the Apache 2.0 license. This move marks a significant step in the democratization of artificial intelligenc…
DeepSeek V4 Launch Set for July: CUDA-Free AI for Crypto Devs
The artificial intelligence landscape is bracing for the official release of DeepSeek-V4, which is expected to debut before the end of July 2026. According to industry reports, the upcoming version has already entered s…